In the case of Child Participant 2 (CP2), the mother was a native German
speaker from Hannover and the father a native Welsh-speaker local to
Bangor. As in the case of CP1, the ‘one parent, one language’ model was
almost invariably adhered to when speaking to the child. In this family,
however, the parents used English between themselves, though both had
some knowledge of the other’s native language as well.
Although both infants lived in areas of North Wales where 81.98% and
80.45% of the populations of their respective villages were Welsh-speaking
(2001 census), English remained a powerful influence in society, in the
media, and in recreation. These infants were also hearing their own parents
and many other individuals using English, and were therefore also acquiring
English as a third language, albeit at a much slower pace since their exposure
to English was less frequent, and mostly indirect. The parents in both fami-
lies were professionals, with all four parents having reached at least degree
level education.
Procedures
The aim was to collect naturalistic speech samples on a bi-weekly basis,
alternating between German and Welsh according to the availability of the
parents. Overall, a total of 28 sessions were recorded in German and a total
of 28 sessions in Welsh. All recording sessions involved the child and two
adults (one of which was a parent) – one a German speaker and one a Welsh
speaker. The parent was the lead interlocutor in any one testing situation,
while the other adult operated the digital video-camera. Such an arrange-
ment reflects more accurately the child’s exposure to the two languages and
does not elicit switching from one language to the other (cf. Genesee, 1989;
Meisel, 1989). (Nevertheless, mixed utterances might occur – cf. Cantone,
2007.) Every effort was made to vary the context of each recording session as
much as possible, including recording meal-time rituals, play-time rituals,
etc. Each recording session lasted for approximately half an hour.
Results
As not all the data were transcribed by the time of publication, for the
present analyses, six recordings (three per language) from each child were
analyzed, covering an age range of 2;4 to 2;10 across the two children.
In what follows, we will first present the patterns of gender marking in
the children’s German utterances followed by their patterns of gender mark-
ing in Welsh. This is then followed by a discussion of word order patterns in
German and then Welsh. In both cases, comparisons are drawn with (near-)
monolingual speakers where appropriate. The chapter concludes with a short
discussion of the findings in relation to cross-linguistic influence.
Cross-linguistic Influence and Patterns of Acquisition 53